Automation of routine clinical chromosome analysis. I. Karyotyping by machine

An automated karyotyping system suitable for widespread use in clinical laboratories is described. The software is implemented on a general-purpose, commercially available image analyzer (Magiscan 2) using TV input from a conventional research microscope with minimal modification. The analysis is automatic, but operator interaction is used to resolve difficulties. Extensive experience with a routine clinical workload shows that the system is robust and easy to use and that its use results in a substantially increased laboratory throughput.     

Automatic classification of chromosomes as part of a routine system for clinical analysis

A procedure for automatic classification of G-banded human chromosomes has been implemented on a semiautomated system for routine clinical analysis. Chromosomes represented by their density profiles are described by so-called weighted density distributions (WDDs) by application of a number of weighting functions and classified by a parametric discriminant analysis. During 16 mo of routine use of the system, 2,794 metaphases (127,925 chromosomes) from amniotic fluid have been karyotyped by the system with an error rate of 8-9%. This corresponds to 4-5 errors per metaphase. These errors can immediately be corrected by the operator on a displayed karyogram with a light pen.     

Automation of routine clinical chromosome analysis. II. Metaphase finding

Metaphase finding is an essential activity in chromosome analysis, and there is much to gain from its automation. This paper describes software for automatic metaphase finding developed for use as part of a routine clinical chromosome analysis system, principally for samples from blood and amniotic fluid. Since the metaphase finding and analysis programs were intended to be used widely in clinical laboratories, cost and portability were important design features. The metaphase finder has been implemented on a moderately priced, general-purpose image analyzer (Magiscan 2), which controls a standard research microscope with motorized stage and focus. Metaphases are detected using fast gray-level processing on whole fields of view, followed by binary processing to produce a figure of merit for each detected object. Clinical experience has shown that this ability to rank detected objects on the basis of their suitability for analysis is a critical feature in determining the usefulness of an automatic metaphase finder.     

Automated metaphase finding: an assessment of the efficiency of the METAFER2 system in a routine mutagenicity assay

The efficiency of the automated metaphase finding system METAFER2 is assessed in a routine mutagenicity assay using an aneuploid rat liver cell line treated with various promutagens. Data sets generated by automated and manual selection of metaphases are compared. It is demonstrated that METAFER2 routinely allows an efficient automatic identification of metaphases not only in lymphocyte preparations, but also in preparation from mammalian cell lines with varying chromosome numbers. Although larger slide areas are required for automated compared to manual metaphase scanning, the automatic system is faster by a factor of about 5. The interactive visual elimination of metaphases of insufficient quality is an easy and fast procedure.METAFER2 allows an unbiased selection of metaphases irrespective of their appearance as homogeneously stained first or harlequin-staines second division cells. Random selection of metaphases is neither influenced by various structural chromosome changes nor by increased frequencies of sister-chromatid exchanges.     

Time-saving in biological dosimetry by using the automatic metaphase finder Metafer2.

The amount of time-saving by using the Metafer2 metaphase finder for routine analysis of radiation-induced chromosome aberrations (biological dosimetry) was determined. Metaphases were prepared by standard methods from cultures of human peripheral blood lymphocytes and stained either with Giemsa or with the FPG method. The metaphase finder was used for detecting metaphases on the microscope slides and for automatically processing the evaluation data. In our laboratory, standardized analysis of 1000 metaphases requires at least 3 working days for cell culturing and slide preparation and 51.5 working hours for cytogenetic analysis. When using the metaphase finder the time required for cytogenetic analysis is reduced to 17.3 working hours (time-saving factor: 51.5/17.3 h = 3.0). In our prolonged method, including more than one scoring of each slide and karyotyping of metaphases with chromosome aberrations, the analysis times for 1000 cells are 132 and 70 working hours, respectively (time saving factor: 132/70 h = 1.9).     

Cytogenetic studies after contact with juvenile blastic leukemias]

Persons with long professionell contact to acute childhood leukemia and first-degree relatives had a higher tendency of chromosome bursting off in the metaphases from cultivated lymphocytes. Since persons with contact and relatives for one thing and controls for another showed significant differences in the degree of hurtibility of metaphases, cultivation ane preparation could not be responsible for themselves. It is discussed whether the contact to childhood leukemia may lead to any reaction between peripheral lymphocytes and an infectious agent. This reaction itself leads to a higher susceptibility of metaphases from peripheral lymphocytes to the injuring laboratory conditions.     

Paint-assisted microdissection-FISH: Rapid and simple mapping of translocation breakpoints in the embryonal rhabdomyosarcoma cell line RD.

BACKGROUND: Spectral karyotyping and multiple fluorophore fluorescence in situ hybridisation (M-FISH) facilitate identification of inter-chromosomal rearrangements, but are of low cytogenetic resolution in mapping translocation breakpoints. Reverse chromosome painting yields increased cytogenetic information but isolation of aberrant chromosomes is technically difficult. We have developed the technique of paint-assisted microdissection FISH (PAM-FISH), which enables microdissection of aberrant chromosomes to be carried out easily and rapidly using relatively simple apparatus. METHODS: A selected chromosome paint is hybridised to abnormal metaphases to label a chromosome of interest, which is then microdissected, amplified, labelled by polymerase chain reaction (PCR), and reverse painted onto extended normal metaphases. RESULTS: PAM-FISH was used to reassess structural chromosomal abnormalities identified by molecular cytogenetics in the rhabdomyosarcoma cell line RD. PAM-FISH improved the analysis of virtually all structural abnormalities, identifying six novel translocations and indicating that seven previously described rearrangements were in fact not present in RD. Accuracy of the breakpoint mapping obtained was confirmed by bacterial artificial chromosome-FISH. CONCLUSIONS: PAM-FISH is ideally suited to analysis of tumour metaphases as it is not affected by poor chromosome morphology. Reagents generated by PAM-FISH are also suitable for other investigations, such as mapping using sequence tagged-site PCR or genomic microarrays. PAM-FISH is technically straightforward and could readily be adopted in a routine cytogenetics laboratory for accurate high-throughput analysis of chromosome breakpoints.     

Technical report: application of the Metafer2 fluorescence scanning system for the analysis of radiation-induced chromosome aberrations measured by FISH-chromosome painting

The Metafer2 fluorescence scanning system was used for routine analysis of radiation-induced exchange aberrations measured by fluorescence in situ hybridisation (FISH) chromosome painting in human peripheral lymphocytes. The system enables a rapid and unbiased fully-automated finding and image acquisition of fluorescently stained metaphase spreads. The chromosome aberration analysis is performed interactively from stored digitised processed gallery images, presented on a screen. Appropriate software image filters are available to further improve these pictures by background correction, noise reduction and fluorescence signal enhancement.Data sets generated by computer-assisted and manual scoring of radiation-induced reciprocal translocations (2B) and total 2B (2B+related 'one-way' types) or complete dicentrics (2A) and total 2A (2A+related 'one-ways') involving painted target chromosomes 2, 3 or 4 were compared and no significant differences were found.A linear-quadratic dose-response curve for total translocations (2B+'one-ways'+complex-derived types) based on computer-assisted analysis of 27,741 metaphases with chromosome 4 painting was compared to a curve obtained earlier for manually scored translocations in a set of target chromosomes 1, 4 and 12. After extrapolation to the whole genome, no significant difference between both curves was found.From our results it can be derived that computer-assisted aberration analysis using the Metafer2 system is a reliable alternative to manual analysis. Since time saving for computer-assisted translocation analysis is about 50% compared to manual scoring, this system is highly promising for a practical application in retrospective biodosimetry of human radiation exposure.     

Karyotype analysis of the euploid cell population of a mouse embryonic stem cell line revealed a high incidence of chromosome abnormalities that varied during culture

It is common knowledge that mouse embryonic stem cell (mESC) lines accumulate chromosomal changes during culture. Despite the wide use of mESCs as a model of early mammalian development and cell differentiation, there is a lack of systematic studies aimed at characterizing their karyological changes during culture. We cultured an mESC line, derived in our laboratory, for a period of 3 months investigating its chromosome complement at different times. About 60% of the metaphases analysed were euploid throughout the culture period but, from passage 13, only 50% of the euploid metaphases had a proper chromosome complement. The remaining 50% showed chromosome abnormalities, mainly gain or loss of entire chromosomes, both within the same passage and among different passages analysed. The very heterogeneous spectrum of abnormalities indicates a high frequency of chromosome mutations that arise continuously during culture. The heterogeneity of the aberrant chromosome constitution of 2n = 40 metaphases, observed at different passages of culture, might be due either to their elimination or to a shift towards the hypoeu- or hypereuploid population of those metaphases that accumulate further chromosome abnormalities. The stability of the frequency of eu-, hypoeu- and hypereuploid populations during culture might, however, be due to the elimination of those cells that carry a high mutational burden. Based on our results, we suggest that karyotype analysis of the euploid cell population of mESC lines is necessary when such lines are used in the production of chimeric mice, for their contribution to the germ line, or when they are differentiated into specific cell types.     

Scanning electron microscopy of the G-banded human karyotype

The chromosome structure of human metaphases was observed in the scanning electron microscope (SEM) after exposure to G-banding techniques for light microscopy (LM). Individual chromosomes showed an inherent specificity of quaternary coiling. Circumferential grooves along the chromatids demarcated the individual gyres of the coils, which were shown to correspond to the LM G-banding pattern. An increased number of quaternary coils was observed in prometaphase chromosomes, which were shown to be correlated with the high resolution LM bands. We propose that the observation of G-bands relies on LM visualization of quaternary structure by accumulation of Giemsa stain between the coils.     

Flow cytometric DNA and cytogenetic studies in human tumors: a comparison and discussion of the differences in modal values obtained by the two methods

The numerical chromosome values in 53 human tumors were determined and compared with the modal DNA values as measured by flow cytometry. In tumors with chromosome counts in the diploid and tetraploid range, the modal DNA values were found to correspond to the modal values based on the chromosome counts. In tumors with chromosome counts in the triploid range, however, the modal DNA values were about 15% higher than expected. In order to explain this difference, the ratio between large and small chromosomes in the karyotyped metaphases was assessed. In addition, the DNA content of individual chromosomes, including markers and minutes, was calculated as a reflection of the DNA content of the whole cell. The ratio of large to small chromosomes did not deviate from the normal ratio found in cells with diploid, triploid, and tetraploid chromosome counts. Neither difficulties in karyotyping nor short-comings in the flow cytometric methodology could be used to explain the discrepancy between the expected and empirical modal DNA values. Some of the chromosomes in triploid tumors may, therefore, contain an increased amount of DNA.     

Chromosomal localization of the mouse prealbumin gene (Ttr) by in situ hybridization.

Prealbumin is a serum protein which plays an important role in plasma transport of retinol and thyroxine. The accumulation of a variant prealbumin is associated with a hereditary disorder, familial amyloidotic polyneuropathy (FAP). In situ hybridization with a mouse prealbumin gene cDNA probe was carried out in mouse fibroblasts. Analysis of 114 R-banded metaphases showed that 13% of the total grains were located on chromosome 4 and 46% of the grains on this chromosome were in the region C6-D1. Linkage and syntenic group analysis showed that the prealbumin gene (Ttr) is located between two syntenic groups on mouse chromosome 4, which corresponded to two syntenic groups present on human chromosomes 1 and 9.     

Distribution of radiation induced lesions in human chromosomes and dose-effect relation analysed with G-banding

Summary Human female lymphocytes were exposed to X-rays in vitro at 7 different doses between 40–280 R. In 830 metaphases chromosome analyses were carried out with either conventional staining or G-banding, respectively. 486 breakpoints are non-randomly distributed between chromosomes and chromosome arms. An excess of lesions was present in chromosomes 1 and 5 or in lp. 85% of the lesions were located in G-negative bands (pale G-bands). 29% of all lesions appeared in either the last terminal pale band (21%) or in the centromere region (8%).With regard to an application of G-banding for a biological dose-estimation, the dose-response relations of dic and ace were analysed. Although G-banding enables detailed analysis of the whole karyotype it cannot be recommended for cytogenetic routine analyses in medical radioprotection monitoring, without suitable automated scoring techniques. Dose estimations based on the frequency of dic and carried out with conventional staining cannot be essentially improved at present with banding. Nevertheless, by banding criteria for a correct evaluation of other aberration types, e.g. ace, can be provided. This is a prerequisite for the calculation of representative dose-effect curves.     

Clinical applications of flow karyotyping in myelocytic leukemia by stimulation of different subpopulations of cells in blood or bone marrow samples

Examples are presented in which normal as well as abnormal chromosome distributions could be obtained from the same individual by means of bivariate flow karyotyping. Selective stimulation of T-lymphocytes obtained by E-rosetting from the blood of a patient with acute myelocytic leukemia resulted in a normal flow karyogram. The specific stimulation of myelocytic leukemia cells with granulocyte-macrophage colony stimulating factor (GM-CSF) and interleukin 3 (IL-3) yielded flow karyograms displaying the leukemia-associated chromosome abnormalities. The resulting flow karyograms could be used to discriminate between homolog differences, which appear normally in virtually every individual, and leukemia-associated chromosomal aberrations. In the case of a female chronic myelocytic leukemia patient who received bone marrow form an HLA-identical male donor, specific stimulation of various subsets of cells enabled to discriminate between leukemic host cells and non-leukemic donor cells. Both the leukemia-specific translocations and sex chromosomes were used as markers to analyse the flow karyograms obtained from the same sample.     

Ring chromosome 21 in the mother and 21/21 translocation in the fetus: karyotype: 45,XX,-21,-21,+t(21;21)(p11;q11)

In 1984 we reported a ring chromosome 21 in a normal woman with recurrent fetal wastage (Kleczkowska and Fryns, 1984). A 46,XY normal fetal karyotype was found after prenatal diagnosis at 14 1/2 weeks in a third pregnancy of this woman. In the present paper we report the prenatal diagnosis of a 21/21 translocation in a female fetus from her fourth pregnancy.     

Fluorescence in situ hybridization on paraffin-embedded abortion material as a means of retrospective chromosome analysis

A fluorescence in situ hybridization (FISH) procedure was used to detect chromosome abnormalities in archival abortion material. Nuclei were isolated from 50-m-thick tissue blocks from 18 selected and karyotyped abortions. Five probes for repetitive centromeric sequences of chromosomes 1, 16, 18, X and Y were used. For each chromosome, at least 200 nuclei were scored blindly, i.e. without knowledge of the karyotype. The FISH results obtained were compatible with the cytogenetic data in 14 cases. There were four discrepancies. Two of these were observed for cases karyotyped as trisomy 16. Furthermore, FISH results showed trisomy 18 in two cases having normal chromosomes 18 and 18q+, respectively. The latter case was not discrepant if the structural rearrangement involved chromosome 18 material. The remaining discrepancies could be explained by chromosomal mosaicism. Admixture of normal maternal cells was also noted. It is concluded that FISH can be used to study retrospectively the presence of chromosome abnormalities in abortion material. However, the quality obtained after the use of fresh material is superior.     

用原位杂交法定位猪乳铁蛋白基因于染色体2q^12

本研究以非放射性标记的猪乳铁蛋白（Ｐｏｒｃｉｎｅ Ｌｉａｃｔｏｆｅｒｒｉｎ简称ＰＬＦ）ｃＤＮＡ为探针,通过染色体原位杂交法,对ＰＬＦ基因了染色体进行了染色体定位。实验中采用金胶抗体技术并结合使用银增强系统,提高了方法的灵敏度。利用染色体的组型分析,对杂交点的分布进行了统计学分析。５２％（２６／５０）的分裂相在第２号染色体具银粒分布,实验结果表明：ＰＬＦ基因定位于猪２号染色体２ｑ＾１２区域。     

Computer-assisted karyotyping with human interaction.

A system for machine assisted karyotyping and chromosome analysis has been developed. The system uses a drum- or TV-scanner as input device, runs provisionally in 32 K memory, and also allows human interaction on several stages. The accuracy with which banded chromosomes are karyotyped depends strongly on the type of classifier and varies from 40 up to 80%. The accuracy of the human assisted classifier (98%) comes close to that of a skilled technician (99.5%) using manual chromosomal analysis. Due to technical and memory limitations, the time necessary for the karyotyping of one cell is too long and depends on the interaction time; however karyotyping within 5 min, including human interaction, will be possible in the near future.     

Multi-locus (ML)-FISH is a reliable tool for nondisjunction studies in human oocytes

In the present study, we developed a fluorescence in situ hybridization (FISH) strategy, which allows a reliable determination of the chromatid number of specific chromosomes in mature human oocytes. 168 unfertilized oocytes were analyzed by dual-color FISH with two direct-labeled locus-specific DNA probes for chromosome 13 and 21. To exclude FISH failures, metaphases with abnormal signal patterns were reanalyzed by multi-locus-FISH (ML-FISH) for chromosome 13 and 21. Following dual-color FISH, abnormal signal patterns were detected in 21 out of 108 metaphases (19.4%). 17 of these metaphases were reanalyzed by ML-FISH. In contrast to the first FISH, seven metaphases showed normal signal patterns after rehybridization, whereas ten metaphases remained abnormal. Out of these real aneuploid metaphases, five showed gain or loss of a single signal (= chromatid), two showed missing double signals (= chromosome) and three showed both. In conclusion, locus-specific FISH probes facilitate differentiation between first meiotic nondisjunction of whole chromosomes and prematurely divided chromatids. Moreover, simultaneous hybridization with a second locus-specific probe on the same chromatid (ML-FISH) helps to differentiate between FISH failures and real meiotic division errors and therefore, allows a more reliable analysis of aneuploidies in human oocytes.     

An analysis of human sperm chromosome breakpoints.

Sperm chromosome analysis of 19 sperm donors with either normal or balanced karyotypes was carried out in order to explore the nature of sperm chromosome structural aberrations. A total of 2,389 cells (range 36-298/donor) were karyotyped after in vitro penetration of hamster eggs. The median percentage of sperm structural aberrations was 9.3% (SD +/- 4.7; range 0%-17.8%), with a total of 247 breakpoints, of which 220 could be characterized fully. Two sets of donors were studied in two different centers: center 1 (United States) and center 2 (Spain). The frequencies of nonrejoined and rejoined chromosome-type aberrations were very similar between center 1 and center 2: 83.6% and 10.0%, and 75.0% and 10.3%, respectively. Chromatid-type aberrations were more frequent in center 2 (14.7%) than in center 1 (6.4%) (P = .037). Chromosome 4 had less than the expected number of breakpoints (P < .001). A positive significant correlation was found between sperm breakpoints reported in this study and sites of balanced chromosome de novo rearrangements detected at prenatal diagnosis and reported in the literature (P = .0001).